videomae-base-finetuned-signlanguage-3

This model is a fine-tuned version of MCG-NJU/videomae-base on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 1.5381
• Accuracy: 0.6667

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

• learning_rate: 1e-05
• train_batch_size: 32
• eval_batch_size: 32
• seed: 42
• optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.1
• training_steps: 7200

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
4.2866	0.0043	31	4.2592	0.0134
4.2806	1.0043	62	4.2450	0.0134
4.2427	2.0043	93	4.2292	0.0134
4.2279	3.0043	124	4.2225	0.0268
4.2165	4.0043	155	4.2176	0.0201
4.225	5.0043	186	4.2126	0.0268
4.2079	6.0043	217	4.2061	0.0134
4.2043	7.0043	248	4.2039	0.0134
4.1945	8.0043	279	4.1983	0.0201
4.1946	9.0043	310	4.1962	0.0134
4.1777	10.0043	341	4.1923	0.0268
4.1941	11.0043	372	4.1824	0.0268
4.1509	12.0043	403	4.1603	0.0201
4.1308	13.0043	434	4.1267	0.0336
4.0462	14.0043	465	4.0901	0.0336
4.0389	15.0043	496	4.0516	0.0537
3.9851	16.0043	527	4.0059	0.0604
3.9384	17.0043	558	3.9735	0.0604
3.8406	18.0043	589	3.9260	0.0671
3.8223	19.0043	620	3.8081	0.1007
3.7223	20.0043	651	3.8393	0.1275
3.5692	21.0043	682	3.6683	0.0872
3.4716	22.0043	713	3.6782	0.1544
3.3267	23.0043	744	3.4783	0.1409
3.1245	24.0043	775	3.2797	0.2349
2.9207	25.0043	806	3.2165	0.2282
2.824	26.0043	837	3.0733	0.3221
2.6562	27.0043	868	3.0572	0.2282
2.5454	28.0043	899	2.8509	0.4094
2.3503	29.0043	930	2.7583	0.3691
2.3168	30.0043	961	2.6929	0.4228
2.1677	31.0043	992	2.5752	0.4631
1.9252	32.0043	1023	2.5122	0.4430
1.9618	33.0043	1054	2.4168	0.5101
1.8355	34.0043	1085	2.3845	0.4966
1.6116	35.0043	1116	2.3630	0.5034
1.4548	36.0043	1147	2.2536	0.4765
1.4548	37.0043	1178	2.1817	0.5168
1.3441	38.0043	1209	2.1902	0.5101
1.2029	39.0043	1240	2.1419	0.4966
1.2918	40.0043	1271	2.0249	0.5638
1.1693	41.0043	1302	2.0413	0.5436
1.1327	42.0043	1333	1.9967	0.5705
1.0055	43.0043	1364	1.9504	0.5705
0.9258	44.0043	1395	1.8976	0.5705
0.9115	45.0043	1426	1.8433	0.6107
0.8194	46.0043	1457	1.9384	0.5638
0.8492	47.0043	1488	1.8381	0.5638
0.7503	48.0043	1519	1.8764	0.5570
0.612	49.0043	1550	1.8364	0.5235
0.5202	50.0043	1581	1.7960	0.5570
0.6069	51.0043	1612	1.8308	0.5570
0.4907	52.0043	1643	1.6718	0.6174
0.537	53.0043	1674	1.7595	0.5705
0.4723	54.0043	1705	1.7004	0.5772
0.4402	55.0043	1736	1.7347	0.5570
0.3488	56.0043	1767	1.7399	0.6040
0.2996	57.0043	1798	1.6006	0.6040
0.3298	58.0043	1829	1.6110	0.5906
0.301	59.0043	1860	1.6797	0.5638
0.2561	60.0043	1891	1.6586	0.5772

Framework versions

• Transformers 4.49.0
• Pytorch 2.0.1+cu118
• Datasets 3.3.2
• Tokenizers 0.21.0